DNABERT1 RefSeq400 Vs SpliceAI 400 on RefSeq400¶
In [1]:
from utility_transformers import plot_predictions_results_comapre, plot_predictions_results
nbins = 100
min_intron_length = 49
max_intron_length = 1100000
min_donor_length = 19
max_donor_length = 11000
min_acceptor_length = 19
max_acceptor_length = 11000
top_n_signals = 10
break_y_min, break_y_max, plotly_histogram_y, ystep = 50000, 50000, 350000, 50000
predictions_path_1 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__RefSeq400/"
negative_predictions_path_1 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__RefSeq400_Negative"
predictions_path_2 = "/data/SpliceUp/prediction/spliceAI_refSeq400/"
negative_predictions_path_2 = "/data/SpliceUp/prediction/spliceAI_refSeq400_Negative/"
output_folder = "/data/SpliceUp/DeepSAP-main/scripts/plots/prediction/SpliceAI_vs_DNABERT1__RefSeq400_RefSeq400/"
pickled_junctions_path = "/data/SpliceUp/pickles/not_strict_annotated_junctions_table__RefSeq__400.pkl"
plot_predictions_results_comapre(pickled_junctions_path, predictions_path_1, negative_predictions_path_1, predictions_path_2, negative_predictions_path_2, output_folder, "DNABERT1 RefSeq400", "SpliceAI 400", "RefSeq400", break_y_min, break_y_max, plotly_histogram_y, ystep, nbins, min_intron_length, max_intron_length, min_donor_length, max_donor_length, min_acceptor_length, max_acceptor_length, top_n_signals, is_show_legend=True)
Plotting scores vs intron length
Plotting scores vs acceptor and donor length
Plotting violins of scores per bitype and signal
DNABERTT MS150 vs DNABERT1 RefSeq150 on RefSeq150¶
In [2]:
from utility_transformers import plot_predictions_results_comapre
nbins = 100
min_intron_length = 49
max_intron_length = 1100000
min_donor_length = 19
max_donor_length = 11000
min_acceptor_length = 19
max_acceptor_length = 11000
top_n_signals = 10
break_y_min, break_y_max, plotly_histogram_y, ystep = 50000, 50000, 350000, 50000
predictions_path_1 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__RefSeq150/"
negative_predictions_path_1 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__RefSeq150_Negative/"
predictions_path_2 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__MultiSpecies150_RefSeq150/"
negative_predictions_path_2 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__MS150__RefSeq150_Negative/"
output_folder = "/data/SpliceUp/DeepSAP-main/scripts/plots/prediction/DNABERT1__MS150_vs_DNABERT1__RefSeq150_RefSeq150/"
pickled_junctions_path = "/data/SpliceUp/pickles/not_strict_annotated_junctions_table__RefSeq__150.pkl"
plot_predictions_results_comapre(pickled_junctions_path, predictions_path_1, negative_predictions_path_1, predictions_path_2, negative_predictions_path_2, output_folder, "DNABERT1 RefSeq150", "DNABERT1 MS150", "RefSeq150", break_y_min, break_y_max, plotly_histogram_y, ystep, nbins, min_intron_length, max_intron_length, min_donor_length, max_donor_length, min_acceptor_length, max_acceptor_length, top_n_signals, is_show_legend=True)
Plotting scores vs intron length
Plotting scores vs acceptor and donor length
Plotting violins of scores per bitype and signal
DNABERTT MS150 vs DNABERT1 PF150 on PF150¶
In [3]:
from utility_transformers import plot_predictions_results_comapre, plot_predictions_results
nbins = 100
min_intron_length = 0
max_intron_length = 110000
min_donor_length = 0
max_donor_length = 11000
min_acceptor_length = 0
max_acceptor_length = 11000
top_n_signals = 10
plotly_histogram_y = 8000
predictions_path_1 = "/data/SpliceUp/prediction/PREDICT__DNABERT1_k6__Splice__PF_r57_150/"
negative_predictions_path_1 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__PF_r57_150_Negative/"
predictions_path_2 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__MultiSpecies150_PF_r57_150/"
negative_predictions_path_2 = "/data/SpliceUp/prediction/PREDICT__DNABERT1__MS150__PF_r57_150_Negative/"
output_folder = "/data/SpliceUp/DeepSAP-main/scripts/plots/prediction/DNABERT1__MS150_vs_DNABERT1__PF150_PF150/"
pickled_junctions_path = "/data/SpliceUp/pickles/not_strict_annotated_junctions_table__PF_r57__150.pkl"
top_n_signals = 10
# break_y_min, break_y_max, plotly_histogram_y = 20000, 1001, 10000
break_y_min, break_y_max, plotly_histogram_y, ystep = 1000, 1000, 10000, 2000
plot_predictions_results_comapre(pickled_junctions_path, predictions_path_1, negative_predictions_path_1, predictions_path_2, negative_predictions_path_2, output_folder, "DNABERT1 Malaria150", "DNABERT1 MS150", "Malria150", break_y_min, break_y_max, plotly_histogram_y, ystep, nbins, min_intron_length, max_intron_length, min_donor_length, max_donor_length, min_acceptor_length, max_acceptor_length, top_n_signals, is_show_legend=True)
# plot_predictions_results_comapre(pickled_junctions_path, predictions_path_1, negative_predictions_path_1, predictions_path_2, negative_predictions_path_2, output_folder, "DNABERT1 Malaria150", "DNABERT1 MS150", "Malria150",plotly_histogram_y, nbins, min_intron_length, max_intron_length, min_donor_length, max_donor_length, min_acceptor_length, max_acceptor_length, top_n_signals, is_show_legend=True, plotly_histogram_y=8000)
Plotting scores vs intron length
Plotting scores vs acceptor and donor length
Plotting violins of scores per bitype and signal
